Lecture Notes

Math 101 Mathematics for Social Sciences

I smail S
 uayip G¸loº
glu

January 2, 2023
2
Contents

1 Numbers 5
Integers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Rational Numbers . . . . . . . . . . . . . . . . . . . . . . . 9
Operations of addition and multiplication on Q . . . . 10
Decimal Expansion of rational numbers . . . . . . . . . 12
Real Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Absolute value function and the distance on R and
intervals(=connected subsets of real axes) . . . 15
Roots and Exponents . . . . . . . . . . . . . . . . . . . 17

2 First Order Equations, Inequalities and Lines in the

Plane 21
Polynomial Functions of degree one and Lines . . . . . . . . 23
Extreme Values of a linear function aX+bY in two variables in

a polygonal region in the plane (special Case of Linear

Programming) . . . . . . . . . . . . . . . . . . . . . . . 26

3 Polynomial Functions of degree two (Quadratic Func-

tions) 29

4 Matrices 33
Matrix Operations . . . . . . . . . . . . . . . . . . . . . . 34
Properties of these Matrix operations . . . . . . . . . . 36
Row Reduction and Row Reduced Echelon Matrices . . . . . 37

5 Linear Systems of Equations and Their Solutions 41

How do we decide whether a given system is consistent,
that is has a solution or not. . . . . . . . . . . . 43
How can we Önd all solutions if we know the system is
consistent? . . . . . . . . . . . . . . . . . . . . . 44
3
4 CONTENTS

6 Di§erentiation 47
Di§erentiation Rules . . . . . . . . . . . . . . . . . . . . . . 49
Applications of the derivative . . . . . . . . . . . . . . . . . 50
 Chapter 1

Numbers

Integers
We shall denote by Z the collection, that is, the set of all integers.
So
n2Z
means that "the object n is an element of the set Z" or, in this
concrete context, "n is an integer".
We have
Z = f::::; 2; 1; 0; 1; 2; :::::g;
7 2 Z
1738 2 Z
p 1
4 2= Z; 2 =Z
2 p
mean that 7 and 1738 are integers but 4(=the number x such
that x2 = 2)and 12 (=the number y such that y  2 = 1) are not integers
.
We have two basic binary operations on the set of integers, namely
the addition and multiplication: For any two integers a and b we have
two other uniquely determined element of Z denoted by a + b and a  b
called the sum and the product of a and b; the assignments of these
to the pair (a; b) are called respectively addition and multiplication:
(a; b) !
7 a+b
(a; b) ! 7 a  b:
5
6 Numbers

Of course one can deÖne many more binary operations on Z for example

(a; b) 7! a  b := a  b + (a) + (b):

But the addition and multiplications are of fundamental importance,

they are basic.
Properties of these operations ( or arithmetical rules to which these
operations obey)

1. Both are associative, that is

(a + b) + c = a + (b + c) and (a  b)  c = a  (b  c)

for any a; b; c 2 Z. This property implies that if you add number

of integers a1 ; a2 ; :::; an successively in a given order, then the sum
does not depend on the way you realize this operation, that is,
the result of your operation does not depend on the way you put
paranthesis in a1 + a2 +    + an or a1  a2      an : For example

a + b + c + d = (a + b) + (c + d) = a + (b + (c + d)) = a + ((b + c) + d)
= (a + (b + c)) + d = ((a + b) + c) + d

similarly for the multiplication

abcd = a(b(cd)) = a((bc)d) = (ab)(cd) = ((ab)c)d = (a(bc))d

2. Both are commutative, that is

a + b = b + a and a  b = b  a

for any a; b 2 Z.

3. With respect to each of these operations there exists an identity

element , that is there exist very speciÖc elements namely Zero, 0
and One, 1 integers playing the role of the identity element with
respect to these operations, that is

a + 0 = a and a  1 = a

for any a 2 Z.

4. Existence of the inverse element with respect to addition : For

any a 2 Z there exists a uniquely determined b 2 Z such that
a + b = 0. This b is called the additive inverse of a or the
inverse of a with respect to addition and is denoted by the sign
a:
Numbers 7

Thus for any a 2 Z there exist a uniquely determined integer a;

the additive inverse of a; such that a + (a) = 0
What is the additive inverse of 5? Additive inverse of 5 is 5;
since 5 + (5) = 0
What is the additive inverse of 7? Additive inverse of 7 is 7;
since (7) + 7 = 0; that is (7) = 7:
What is the additive inverse of 0? Additive inverse of 0 is 0; since
0 + 0 = 0 that is (0) = 0.

5. If a 2 Z then an element b 2 Z is called a multiplicative inverse

of a if a  b = 1: Only 1 and 1 have multiplicative inverses.

6. Multiplication is distributive over addition, that is

a  (b + c) = (a  b) + (a  c)

for any a; b; c 2 Z.
Furthermore

7. Since a  0 = a  (0 + 0) = (a  0) + (a  0) we get

(a  0) = (a  0) + 0 = (a  0) + ((a  0) + ((a  0))) =

= ((a  0) + (a  0)) + ((a  0)) = (a  0) + ((a  0)) = 0

that is
a0=0
for any a 2 Z. Observe that as we have seen above this property
of 0 is a logical consequence of the above mentioned properties of
these operations.

8. There are no zero-divisors in Z that is the equation a  b = 0

implies that either a = 0 or b = 0:
Therefore cancellation is possible for nonzero factors, that is,

a  x = a  y and a 6= 0 implies that x = y:

Obviously cancellation for addition is always possible, that

is
a + x = a + y implies x = y for any a:

9. For any a 2 Z the additive inverse of the additive inverse of a is

a; that is
(a) = a
8 Numbers

because this follows from the following equations by cancellation

a + (a) = 0
(a) + ((a)) = 0=

Also we have
(1)  a = (a)
for any a 2 Z because a+(1)a = 1a+(1)a = (1+(1))a =
0  a = 0 thus (1)  a must be the additive inverse of a:

10. A new operation

subtraction (a; b) 7! a  b := a + (b)

for any a; b 2 Z .This operationis not associative and is not com-

mutative:

((2)  (3))  (1) = (5)  (1) = 4

(2)  ((3)  (1)) = (2)  (4) = 6
(2)  (3) = 5 but (3)  (2) = 5:

11. We have on Z a division algorithm which says that for any

a; b 2 Z with b > 0 there exist uniquely determined integers q
and r such that a = qb + r where 0  r < b . q is called the
quotient and r the remainder obtained when a is divided by
b:
1352 = 17  79 + 9 thus we Önd the remainder 9 if we divide 1352
by 17, the quotient is 79,
if we divide 1352 by 17 we Önd the quotient 80 and the re-
mainder 8 since 1352 = 17(80) + 8:

12. Representation of numbers to the base 10 and how to apply the

arithmetic operations of addition , subtraction, multiplication
and the division algorithm of numbers represented to the base
10:
We write usually our numbers using ten digits (symbols) 0; 1; 2; 3; 4; 5; 6; 7; 8; 9
to the base 10 :
The integer "three tausend fourty seven" we write using "base
10" notation as 3047 meaning 3  103 + 0  102 + 4  10 + 7:
We can express every positive integer n 2 Z, n  0 to the base 10
as n = am am1    a2 a1 a0 meaning that n = a0 + a1 10 + a2 102 +
a3 103 +    am 10m where ak 2 f0; 1; 2; 3; 4; 5; 6; 7; 8; 9g
Numbers 9

To Önd this representation you have to divide n by 10 and get

n = q0 10 + r0 with 0  r0 < 10 and put a0 = r0 : Then divide
n  r0
q0 = 2 Z by 10 and get q0 = q1 10 + r1 with 0  r1 < 10
10
and put a1 = r1 then divide q1 by 10 and .... go on in this way.
Arithmetic operations How do we add and multiply integers?
(This you know !!??)
Long division: Given a = 845072 and b = 23: Find q and r if
possible such that a = qb + r and 0  r < b :
845072
23
=?? Write (to the base 10) with digits a1 a2 a3 a4 a5 a6
(a1 = 8; a2 = 4; a3 = 5; a4 = 0; a5 = 7; a6 = 2)

Step 1 We look for b = 23 in a1 = 8 : q1 = 0 and r1 = 8

Step 2 We look for b = 23 in 84 = 10  r1 + a2 : q2 = 3and r2 =
84  23  3 = 15
Step 3 We look for b = 23 in 155 = 10  r2 + a3 : q3 = 6 and
r3 = 155  6  23 = 17
Step 4 We look for b = 23 in 170 = 10  r3 + a4 : q4 = 7 and
r4 = 170  7  23 = 9
Step 5 We look for b = 23 in 97 = 10  r4 + a5 : q5 = 4 and
r5 = 97  4  24 = 5
Step 6 We look for b = 23 in 52 = 10  r5 + a6 : q6 = 2 and
r6 = 52  2  23 = 6
Result : q = q1 q2 q3 q4 q5 q6 = 036742 = 36472 and r = r6 = 6:
Check : q  b + r = 36742  23 + 6 = 845 072

Rational Numbers
The need to divide,forces us to enlarge our system of numbers so that
inverses for nonzero elements have multiplicative inverses for nonzero
elements.
a
The symbol (read :a over b) where a; b 2 Z and b 6= 0 is called a
b
fraction of integers and we denote the set of all such fractions by Q,
that is
a
Q = f : a; b 2 Z,b 6= 0g
b
10 Numbers

and call it the set of rational numbers. Here we consider two di§erent
a c
symbols and as representing the same object, the same rational
b d
number, if and only if
ad = bc
5 15 0 0
For example = ; = :
12 36 3 7
3x + 4 2
Exercise 1 Find x 2 Z such that = :
6 4x + 3
3x + 4 2
Solution = means that (3x + 4)(4x + 3) = 2  6 that
6 4x + 3
is ((3x + 4)  4x) + ((3x + 4)  3) = (3x)  (4x) + 4  (4x) + (3x)  3 + 4  3 =
12x2 + 16x + 9x + 12 = 12x2 + 25x + 12 = 12
Therefore 12x2 +25x = 0 and (12x+25)x = 0: So we see that either
x = 0 or 12x + 25 = 0: As there does not exist any integer satisfying
the second equation we Önd the only solution x = 0:

Operations of addition and multiplication on Q

For any two rational numbers x; y 2 Q we deÖne their sum and product
a c
as follows: there exist fractions ab and dc such that x = and y = we
b d
let
ad + bc
x  y :=
bd
and
ac
xy = :
bd
in this way we get two well deÖned operations on the set of rational
numbers. Well deÖned because the expressions on the right hand sides
of these equations are fractions (observe that ac; bd; ad + bc are integers
and bd 6= 0) and are independent of the representatives of x and y that
a0 c0
is if x = 0 and y = 0 then we can, using ab0 = a0 b and cd0 = c0 d;
b d
check that
ad + bc a0 d0 + b0 c0
=
bd b0 d0
and
ac a0 c 0
= 0 0:
bd bd
We can also deduce that these operations have the following prop-
erties:
1. Both are associative
(a  b)  c = a  (b  c) and (a  b)  c = a  (b  c)
for all a; b; c 2 Q.
Numbers 11

2. Both are commutative

a  b = b  a and a  b = b  a

for all a; b 2 Q.
3. With respect to both operations there exists identity elements 0Q
and 1Q such that x  0Q = x and x  1Q = x for all x 2 Q namely
0 1
0Q = and 1Q = :
1 1
4. With respect to addition every rational number x has an additive
(a) a
inverse x, namely x = if x = :
b b
With respect to multiplication every rational number x 6= 0Q has
a
a multiplicative inverse x1 : if x = 6= 0Q that is a 6= 0 then
b
1 b
x = :
a
5. Multiplication is distributive over addition.
6. We can identify the integers with some rational numbers and
a
imbed Z into Q: a 7! ; a 2 Z then
1
a+b a b
(a + b) 7! = 
1 1 1
and
ab a b
a  b 7!= 
1 1 1
and we see that the operations in Q restricted to the subset Z
coincides with the addition and multiplication of integers.
So from now on we shall write a1 = a for any a 2 Z and use
the same simple notation for the addition and multiplication in
Z and Q, that is
a c ad+bc ad + bc a c ac ac
+ = = and  = =
b d bd bd b d bd bd
and
b ac + b b ab
a+ = and a  = :
c c c c
Furthermore the equation a = qb + r with 0  r < b can and will
be written as
a r
=q+
b b
and q will be called the integer part of the rational number
a r a
and will be called the fractional part of
b b b
12 Numbers

a c
7. Subtraction : ; 2 Q the di§erence of these rational num-
b d
bers in this order
a c a c ad  bc
 := + ( ) =
b d b d bd
a c c
and division by non-zero elements : ; 2 Q and 6= 0;
b d d
the quotient of these rational numbers in this order is

a c a c a d ad
: :=  ( )1 =  = ;
b d b d b c bc
a
a c ad ad
we sometimes write : = cb = = ; as special cases we
b d d
bc bc
a
b a ac a a
have a : = = and : c = b = :
c b b b c bc
c

Decimal Expansion of rational numbers

Similar to the method of writing integers to the base 10 we can repre-
sent rational numbers also to the base 10 : for example
n = 17 1 = 0  7 + 1 n = 0:
1  10 = 1  7 + 3 n = 0:1
3  10 = 4  7 + 2 n = 0:14
2  10 = 2  7 + 6 n = 0:142
6  10 = 8  7 + 4 n = 0:1428
4  10 = 5  7 + 5 n = 0:14285
5  10 = 7  7 + 1 n = 0:142857
showing that n = 71 = 0 + 1  10
1 1 2
+ 4  ( 10 ) + 2  103 + 8  104 +
5 6 1 6
5  10 + 7  10 + 7 10 and we write it to the base 10 in the
form of an inÖnite but repeating (periodic) decimal expansion as n =
0; 142857142857142857142857142857::::::: = 0; 142857

Every rational number is an ultimately repeating (peri-

odic) decimal expansion:
p
n= = [n] + 0; b1 b2 :::bm a1 a2 ::::ar a1 a2 ::::ar a1 a2 ::::ar a1 a2 ::::ar ::::: =
q
[n] + 0; b1 b2 :::bm a1 a2 ::::ar where [n]

is the integer of the rational number n and b1 ; b2 ; ::; bm ; a1 ; a2 ; ::; ar

are digits.

Exercise 2 1. Write 1; 423 as a fraction!

Numbers 13

Let x = 1; 423: Then x is a rational number because it is an

ultimately periodoc decimal expansion. We have 103 x = 1423; 23
and 10x = 14; 23:That is 103 x10x = 990x = 1423; 2314; 23 =
1409
1423  14 = 1409. Thus x = :
990
2. Compute 0; 13 + 0; 307: and 0; 83 + 0; 307:
Let x = 0; 83 + 0; 307 = 0; 838383 + 0; 307307 ) 106 x = 838383 +
307307 + x ) 999999x = 838383 + 307307 = 1145690 .
1145 690 12590
Thus x = = = 1; 145691:
999999 10989

Decimal expansion is not always unique : For example 1; 0 =

0; 9 = 1:

Real Numbers
(In this section I should add some pictures, unfortunately they are
missing for the moment, I hope you can complete these missing pictures
yourself, following the process I tried to describe)
Let us take a line and choose an arbitrary point on it and give
the "name" 0 to this point. The point 0 determines two half-lines,
choose one of them as the "positive" half-line and choose on this half-
line another point and give the "name" 1 to this point. The distance
between these two points we take as our unit length and using this line
segment (or this unit length we determine the points on the positive half
line the points corresponding to positive integers and then determine
their symmetric images with respect to the point 0: In this way we
get the points on the line corresponding to integers. If n is a positive
integer then the point on the positive half-line which has distance n
units from the point 0 is "named " the point n; and if n is a negative
integer then the point on the negative half-line (the half-line which does
not contain the point 1) which has distance n units from the point 0
is "named " the point n:
If x is a rational number with integer part n and the fractional part
a a
so that x = n+ where 0 < a < b then we can locate x using Thalesí
b b
theorem as the point between the points n and n + 1 corresponding to
a
as follows : Draw an arbitrary line passing through the point n and
b
take one of the haláines determine by the point n and take the points
Pa and Pb on this chosen half-line which have integer distances a and
b respectively from the point n . Then draw the line joining Pb to the
14 Numbers

point n + 1 and take the line l parallel to this line and passing through
Pa : The intersection of l with the number line we call the number x:
Observe that between any two rational numbers there are inÖnitely
many rational numbers, set of rational numbers is dense on the line, but
there are gaps that is there are points on the line which are not named
by any rational number. For example, the length d of the diagonal
of a square with side length 1 satisÖes by Pythagoras the equationp
12 + 12 = d2 ; that is d2 = 2: that number is denoted by d = 2 and
a
is not rational , it is not possible to write d = for positive integers a
b
and b:
a
Now we sketch the proof of this fact : Suppose that d = for some
b
integers a and b in the simplest form, that is, a and b do not have
a common factor so that further cancellation is not possible. Then
a2 = 2b2 :If a would be an odd integer than its square would also be
odd. But a2 is even , therfore a must be even, that is, a = 2c for some
integer c , Then we have 4c2 = 2b2 and hence b2 = 2c2 : This shows that
b is also even/ So both a and b are divisible by 2 which contradicts our
assumption about a and p b:
We can represent 2 with a decimal expansion:
Compute the decimal expansion of the square root of a
number x (for example x = 77333 963
)
77 963
Step 0 Write x is a decimal expansion. For example x = =
333
234; 123::: then write it to the base 102 : x = 02102 +34+12 1012 +
31  1014 + 23  1016 + :::: = 234; 123123123 = Y1 Y2 ; Y3 Y4 Y5 Y6 ::::with
digits
Y1 = 02; Y2 = 34; Y3 = 12; Y4 = 31; Y5 = 23; :::::
Put N0 = 0 , M0 = 0 and R0 = 0:
Step 1 Let N1 = 102 R0 + Y1 = 2 and Önd the largest integer u1 such
that u21  N1 : u1 = 1 and R1 = N1  (M0 + u1 )u1 = 2  1 = 1
and M1 = 10(M0 + 2u1 ) = 20
Step 2 Let N2 = 102 R1 +Y2 = 134 and Önd the largest number u2 such
that [M1 +u2 ]u2  N2 that is u2 = 5 and R2 = N2 (M1 +u2 )u2 =
134  125 = 9 and M2 = 10(M1 + 2u2 ) = 300
Step 3 Let N3 = 102 R2 + Y3 = 912 and Önd the largest number u3
such that (M2 + u3 )u3  N3 that is u3 = 3 and R3 = N3 
(M2 + u3 )u3 = 3 and M3 = 10(M2 + 2u3 ) = 3060
Step 4 Let N4 = 102 R3 + Y4 = 331 and Önd the largest number u4
such that (M3 + u4 )u4  N4 that is u4 = 0 and R4 = N4 
(M3 + u4 )u4 = 331 and M4 = 10(M3 + 2u4 ) = 3060
Numbers 15

Step 5 Let N5 = 102 R4 + Y5 = 3333 and Önd the largest number u5

such that (M4 + u5 )u5  N5 that is u5 = 1 and R5 = N5 
(M4 + u5 )u5 = 272 and M5 = 10(M4 + 2u5 ) = 30620
r
77 963
Go on in this way Result : = u1 u2 ; u3 u4 u5 ::::: = 15:301::::
333
Check : 15:3012 = 234:12

p p p
Exercise 3 Find the decimal expansions of 2; 3 and 27; 534 up
to 3 decimal places after comma
Real numbers all the possible decimal expansions and Öll
the real axis completely.

Absolute value function and the distance on R and

intervals(=connected subsets of real axes)
Let us deÖne the absolute value of a real number a by

a if a  0
jaj := :
a if a < 0
For example j3j = 3; j2j = 2; j0j = 0:
We have the following properties:
1. jxj  0 for any x 2 R and jxj = 0 only if x = 0:
2. jx  yj = jxj  jyj for any x; y 2 R:
3. jx + yj  jxj + jyj for any x; y 2 R:
Exercise
1. Find all real numbers x satisfying j4  xj = x:
Solution Since j4  xj is nonnegative for any x 2 R we see that
x  0: Therefore we have 4  x  4 > 0 and hence j4  xj =
4  x = x This gives 4 = 0 which is never true. So there exists
no real number such that j4  xj = x:
2. Find all real numbers x satisfying j4 + xj = x:
Solution 4 + x  0 if x  4 and 4 + x < 0 if x < 4: Therefore
j4 + xj = 4 + x if x  4 and j4 + xj = (4 + x) if x < 4: Thus
we get 4 + x = x if x  4 and (4 + x) = x if x < 4: In
the Örst case we get x = 2 and no solution in the second case
So there is only one real number satisfying j4 + xj = x; namely
x = 2:
16 Numbers

3. Find all real num.bers x satisfying j40 + 3xj  jx + 12j + 5:

Solution...........

We deÖne the distance between two real numbers x and y by

dist(x; y) := jx  yj :
Clearly dist(x; y)  0 and dist(x; y) = 0 if and only if x = y: Fur-
thermore dist(x; y) = dist(y; x) and dist(x; z)  dist(x; y) + dist(y; z):

Exercise 4 1. Find the set of all x 2 R such that jx  5j  3:

The following subsets of the set of real numbers are called intervals.
Let a and b be real numbers with a < b (that is positive di§erence
b  a) then we use the following notation :
[a; b] := fx 2 R : a  x and x  bg
[a; b) := fx 2 R : a  x and x < bg
(a; b] := fx 2 R : a < x and x  bg
(a; b) := fx 2 R : a < x and x < bg
(a; 1) := fx 2 R : a < xg
[a; 1) := fx 2 R : a  xg
(1; b) := fx 2 R : x < bg
(1; b] := fx 2 R : x  bg
(1; 1) := R

The midpoint of the interval [a; b] is the point m with ma = bm
a+b
that m = : Then distance of the midpoint from the boundary
2
points a and b is the radius r of the interval
a+b a+b ba
r = jm  aj =  a = jm  bj = b  =
2 2 2
and it is of course half of the length of the interval which is ba: Then
the points of the intervals are the points which have distance from the
midpoint less than or equal to the radius
 
 a + b  ba
[a; b] = fx 2 R : x   g:
2  2

For example  
 3 + 5  5  (3)

[3; 5] = fx 2 R : x   g = fx 2 R : jx  1j 
2  2
4g;  
 7 
(7; 1) = fx 2 R : x 
+ (1)  < 1  (7) g = fx 2 R :
2  2
jx  4j  3g;
Numbers 17

Exercise 5 Express the set fx 2 R : jx  1j > 2g as a union of

intervals . This the set of all real numbers which are not lying in
the set fx 2 R : jx  1j  2g = [1  2; 1 + 2] = [1; 3]: Therefore
fx 2 R : jx  1j > 2g = R [1; 3] = (1; 1) [ (3; 1):

Roots and Exponents

For a 2 R and n 2 Z; n > 0 we deÖne

an := a  a      a

where we have on the right hand side exactly n factors. (Read an as "
a to the power n " or simply " a to n " .) a is the base and n is the
exponent.
For example (2)5 = 32; 73 = 49  7 = 343; 0n = 0; 1n = 1:
Clearly we have

an  am = an+m
(an )m = anm
an  bn = (a  b)n

for any real numbers a and b and any positive integers n and m:

Remark 6 We have almost never (a + b)n = an + bn : *when is it true

?*
We have the so-called binomial formula

n(n  1) n2 2 n(n  1)(n  2) n3 3

(a + b)n = an + nan1 b + a b + a b + 
2 23
n(n  1)(n  2) 3 n3 n(n  1) 2 n2
+ ab + ab + nabn1 + bn :
23 2
How can we extend taking powers to negative exponents?
If a 6= 0 then a has a multiplicative inverse, we had denoted it by
1
a but did not consider it as "a to the power (1) " but only as the
multiplicative inverse of a: Now using it we want to deÖne the powers
of a nonzero real number to negative exponents: If 0 6= a 2 R and m
is a negative integer (then m is positive) and we deÖne

1 1
am = (a1 )(m) = ( )m = m = (am )1
a a
and add to it also
a0 = 1:
Then we have still the same formulas
18 Numbers

an  am = an+m
(an )m = anm
an  bn = (a  b)n

for any a; b 2 R  f0g and any n; m 2 Z; clearly together with

an
= anm :
am

Example 7 24 + 3  42 = 214 + 3  412 = 16 1

+ 3  161 4
= 16 = 14 (= 22 )
or 24 + 3  42 = 24 + 3  (22 )2 = 24 + 3  22(2) = 24 + 3  24 =
(1 + 3)24 = 4  24 = 22  24 = 22+(4) = 22 :

1 1+23
+1 9 32 1
Example 8 (23 + 1)35 = 23
35
= 23
35
= 23 35
= 23 35
= 23 352
=
1 1 1 3
23 3 3
= (23) 3 = 63 = 6 :

For a 2 R and n 2 Z; n > 0 we look at the the equation

xn = a:

If a = 0 then it has only one solution namely x = 0: Assume that

a 6= 0: Then the equation has

1. no solution x 2 R if n is an even integer and a < 0 (because

y 2  0 for any y 2 R)

2. has two solutions x andpx if n is even and a > 0 , the positive

solution is denoted by n a and
p p
3. exactly one solution if n is odd, which is denoted by n a, and n a
is positive if a is positive and is negative if a is negative.

p
4
p
3
Exercise 9 Find, if possible, 16  27 and Önd all solutions of
(x2 + 1)3 = 1000:

We have for all positive real numbers a 2 R and any rational

n
number r = (n; m 2 Z) we let
m
p p
ar := m an = ( m a)n :
Numbers 19

This formulas should not be applied to negative real numbers and ra-
tional exponents also but may be a reason for serious mistakes:
2 p p
2 = (2)1 = (2) 2 = 2 (2)2 = 4 = 2
2

2 p p
(2) 2 = ( 2 (2))2 canít be computed since 2 is not a real number.

Let us repeat once more for any positive real numbers a and b and
any rational numbers r and s we have

ar  as = ar+s
(ar )s = ars
ar  br = (a  b)r :
20 Numbers
 Chapter 2

First Order Equations, In-

equalities and Lines in the
Plane

DeÖnition 10 A real valued function f of one real variable deÖned on

the subset A  R is a rule which assigns to each element x 2 A a
uniquely determined element y 2 R .We denote this situation by

f : A ! R; x 7! y:

The element y is denoted by f (x) and is called the value of f at x: A

is called the domain of f:The graph of f is the subset f(x; y) : x 2 A
and y = f (x)g of R2 :


2x + 4 if  1  x < 3
Example 11 Let f : [0; 5] ! R be given by x 7! f (x) = 2
3
x4 if 3  x < 8
The graph of f is

2x + 4 if  1  x < 3
f (x) = 2
3
x4 if 3  x < 8
21
22 First Order Equations, Inequalities and Lines in the Plane

6
y
5

-1 1 2 3 4 5 6 7 8
x
-1

-2

Answer the following questions

1. Find [f (0) + f ( 23 )  f (3)]f (5):

2. Find the domain and range of f:
3. Express the following sets as unions of intervals

(a) f x : f (x) = 3g/

(b) f x : f (x) = 0g
(c) f x : f (x) < 3g
(d) f x : f (x)  1g:

Answers (obtained by looking(!) at the graph)

1. [f (0) + f ( 23 )  f (3)]f (5)) = [4 + ( 43 + 4) + (2)]( 10

3
 4) =
22 2 44
3
( 3 ) =  9
2. We see that the domain of f (that is the projection of the graph
on the x-axis) is [1; 8] and the range of f (;the projection of
the graph on the y-axis) is [2; 6]:

(a) f x : f (x) = 3g = ;; that is, f does not take the value 3:
(b) f x : f (x) = 0g) = f2; 6g
(c) f x : f (x) < 3g = ( 12 ; 8]
(d) f x : f (x)  1g = [1; 52 ] [ [ 92 ; 8]
First Order Equations, Inequalities and Lines in the Plane 23

Polynomial Functions of degree one

and Lines
Functions f : R ! R deÖned by the rule x 7! f (x) = a0 xn + a1 xn1 +
   + a1 x + an , where ak ; k = 0; 1; 2; :::; n are Öxed constants and
a0 6= 0 are called real polynomial functions of degree n: In particular
real polynomial functions of one real variable of degree one are functions
of the form
x 7! ax + b; x 2 R

for some Öxed a; b 2 R with a 6= 0: (If a = 0 then the resulting function

is a constant function taking the value b at any x 2 R; if b 6= 0 = a the
function is a polynomial function of degree 0. The constant function
taking the value zero is also considered as a polynomial function but
we do not assign a degree to it.)
Let f : R ! R; x 7! ax + b be a polynomial function of degree
1, a 6= 0. Let us consider two di§erent points on the graph of this
function P1 = (x1 ; y1 ) and P2 = (x2 ; y2 ) with x1 6= x2 : Clearly yk =
f (xk ); k = 1; 2, since these points are on the graph of f: The quotient
of di§erences
y2  y1
=a
x2  x1

is a constant independent of the choice of the points P1 and P2 since

y2  y1 f (x2 )  f (x1 ) (ax2 + b)  (ax1 + b) a(x2  x1 )

= = = = a:
x2  x1 x2  x1 x2  x1 x2  x1

Conversely if g : R ! R is a function for which the di§erence

quotient g(xx22)g(x
x1
1)
= m is a constant for any choice of the points
Pk = (xk ; g(xk )) on its graph then we see that g is a polynomial function
of degree less than or equal to 1.This we see as follows : we Öx a point
P0 = (x0 ; g(x0 )) on the graph and let (x; g(x)) change arbitrarily change
on the graph of the function g then

g(x)  g(x0 )
= m; that is y = g(x) = mx + g(x0 )  mx0 :
x  x0

Therefore the graph of a polynomial function y = f (x) = ax + b is

a line with slope a passing through the point (0; b):
24 First Order Equations, Inequalities and Lines in the Plane

5
y
4

-3 -2 -1 1 2 3
-1 x
-2

-3

-4

-5

-6

The linesy = 1; y = 12 x; y = 12 x  1; y = 2x  3

2

A (straight) line in the plane is the set of points (x; y) satisfying a

linear equation, an equation of the form

ax + by = c

w¨th some f¨xed real numbers a; b; c such that a2 + b2 6= 0:

1. The lines described by the equations ax+by = c and a0 x+b0 y = c0

are equal to each other if and only if there exist a real number
r 6= 0 such that a = ra0 ; b = rb0 ; c = rc0 :

2. The line described by the equations ax + 0y = c (a 6= 0) is a line

c
parallel to the y-axis and passing through the point ( ; 0) on the
a
x-axis (it is not the graph of a function).

3. The line described by the equations ax + by = c (b 6= 0) is the

graph of the Örst degree polynomial function y = mx + n with
m =  ab and n = cb :

4. Two lines given by the equations a1 x+b1 y = c1 and a2 x+b2 y = c2

are

(a) parallel to each other if and only if (their slopes are equal
if and only if) a1 b2 = a2 b1 ;
(b) perpendicular (orthogonal) to each other if and only if
(the product of their slopes is equal to 1 if and only if)
a1 a2 + b1 b2 = 0:
First Order Equations, Inequalities and Lines in the Plane 25

The equation of a line passing through the point P0 = (x0 ; y0 ) and

yy0 y1 y0
1. the point P1 = (x1 ; y1 ) : xx0
= x1 x0

yy0
2. having slope m : xx0
= m;

3. is parallel to the line with the equation ax + by = c : ax + by =

ax0 + by0
yy0
4. is parallel to the line with slope m : xx 0
=m

5. is perpendicular to the line with the equation ax + by = c :

bx + ay = bx0 + ay0
yy0
6. is perpendicular to the line with slope m : xx0
=  m1 :

The distance of the point P0 = (x0 ; y0 ) from the line ` with

the equation ax + by = c : The line passing through P0 and perpen-
dicular to ` has the equation bx + ay = bx0 + ay0 and intersects the
line ` at the point P1 = (x1 ; y1 ) with

ax1 + by1 = c
bx1 + ay1 = bx0 + ay0

which yields that

(a2 + b2 )x1 = ac + b2 x0  aby0

(a2 + b2 )y1 = bc + a2 y0  abx0 :

The distance of the point P0 = (x0 ; y0 ) from the line ` is equal to the
distance d of P0 from P1 . So we have

d2 = (x1  x0 )2 + (y1  y2 )2
ac + b2 x0  aby0  (a2 + b2 )x0 2 bc + a2 y0  abx0  (a2 + b2 )y0 2
= ( ) + ( )
a2 + b 2 a2 + b 2
a2 (c  ax0  by0 )2 + b2 (c  ax0  by0 )2
=
(a2 + b2 )2
(ax0 + by0  c)2
=
a2 + b 2
So d = dist(P0 ; `) =the distance of the point P0 = (x0 ; y0 ) from the
line ` with the equation ax + by = c is

jax0 + by0  cj
d= p :
a2 + b 2
26 First Order Equations, Inequalities and Lines in the Plane

Intersection point of the lines a1 x + b1 y = c1 and a2 x + b2 y = c2

if they are not parallel, that is a1 b2  a2 b1 6= 0 :
We need to Önd (x; y) which satisfy both of these equations. But
such a point is also a solution of the ststem consisting of the equations
b2 (a1 x + b1 y) = b2 c1 and b1 (a2 x + b2 y) = b1 c2 and hence also the
di§erence of these

(a1 b2  a2 b1 )x = b2 c1  b1 c2 :

Similarly we obtain

(a1 b2  a2 b1 )y = a1 c2  a2 c1 :

Extreme Values of a linear function

aX+bY in two variables in a polyg-
onal region in the plane (special
Case of Linear Programming)

Regions satisfying linear inequalities

Let ` be a line and P0 = (x0 ; y0 ) be a point not contained in `: Then
there exist a unique plane  containing ` and P0 :The line ` determines
two closed half-planes 0 and 1 ;only one of them contains the point
P0 ; let us denote that one by 0 : If ` is given by the equation ax+by = c
then these half-planes can be given by the inequalities ax + by  c and
ax + by  c: If ax0 + by0  c then 0 = f(x; y) : ax + by  cg and
1 = f(x; y) : ax + by  cg and if ax0 + by0  c then 0 = f(x; y) :
ax + by  cg and 1 = f(x; y) : ax + by  cg.

Exercise 12 1. Sketch the region bounded by the lines 3x + 2y 

6; 3x + 5y  15 and y  5

2. Express the triangular region with vertices (5; 0); (5; 6) and (2; 0)
in terms of some linear inequalities.

3. Express some of the convex regions in the following Ögure by

means of some linear inequalities and Önd the corners (vertices)
of these regions
First Order Equations, Inequalities and Lines in the Plane 27

y 10
8
6
4
2

-4 -2 2 4
-2 x
-4
-6
-8
-10
-12
-14
28 First Order Equations, Inequalities and Lines in the Plane
 Chapter 3

Polynomial Functions of de-

gree two (Quadratic Func-
tions)

A real polynomial function of one real variable of degree two are func-
tions of the form
x 7! ax2 + bx + c; x 2 R
for some Öxed a; b; c 2 R with a 6= 0:

ax2 + bx + c = 0
is called a quadratic equation. By completing to squares we can
write
b b b c b b2  4ac
ax2 + bx + c = a(x2 + x + ( )2  ( )2 + = a[(x + )2  ]
a 2a 2a a 2a 4a2
Putting  = b2  4ac we see that the equation ax2 + bx + c = 0 has
two distinct real roots
p p
b +  b  
x1 = and x2 =
2a 2a
if  > 0 and one real root with multiplicity two

b
x1 = x2 = 
2a
29
30 Polynomial Functions of degree two (Quadratic Functions)

if  = 0 and has no real roots if  < 0 (In the last case we say that
we have two complex conjugate nonreal roots). In the Örst two cases
we write
ax2 + bx + c = a(x  x1 )(x  x2 )

Now if we look at the function y = f (x) = ax2 + bx + c = a[(x +

b 2 2 4ac
2a
)  b 4a 2 ] then we see that

2
b b +4ac
1. the point P0 = ( 2a ; 4a ) is an extreme point of the graph of
the function. It is the maximum point if a is negative and it is
a minimum point if a is positive,

b b
2. the function takes the same value at the points  2a +t and  2a t
for any t 2 R and this means that the graph of the function is
b
symmetric with respect to the vertical line x =  2a ;

3. if we let the di§erence quotient of the function f at x0 to be

a(t2 x20 )+b(tx0 )
f = f (t)f (x0 )
tx0
then f = tx0
= a(t + x0 ) + b and the
value of this function of t at x0 is equal to the slope of the tangent
line to the graph of the function f at the point (x0 ; f (x0 )); that
is, the equation of the tangent line to the graph of f at (x0 ; f (x0 ))
is
y = f (x0 ) + (2ax0 + b)(x  x0 );

in particular the graph has a tangent line parallel to the x-axis

only at the point P0 :

4. The graph of f

p
b+ 
(a) intersectspthe x-axis at two di§erent points x1 = 2a
and
x2 = b2a  if  = b2  4ac is positive
(b) touches the x-axis at one point (that is , is tangent to the
x-axis) x = b
2a
;if  = 0;
(c) does not intersect the x-axis if  is negative.

Compare the following graphs

1. y = x2
Polynomial Functions of degree two (Quadratic Functions) 31

4
y

-2 -1 0 1 2
x

2. The functions y = x2 ; y = 2x2 ; y = 4x2 ; y = x2

10
y
8

-4 -2 2 4
x
-2

-4

3. The functions y = x2 ; y = (x  3)2 ; y = (x + 2)2

32 Polynomial Functions of degree two (Quadratic Functions)

10
y
9

-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
x

4. The functions y = (x  2)2  4; y = 2(x + 1)2 + 6; y = 12 x2  2

y
6

-4 -2 2 4 6
x
-2

-4

Exercise 13 Sketch the graph of the functions y = (x + 2)2 + 1; y =

(x  2)2 + 4; y = 3x2  12x  4

Exercise 14 Find the points of intersection of the curves in the above

Ögure.
 Chapter 4

Matrices

DeÖnition 15 A real matrix A of size n  m is a rectangular table

with n rows and m columns, each row containing m boxes and each
column containing n boxes , so that the box in the i-th row and j-th
column contains an entry which is a real number and is denoted by
ai;j . We denote the set of all n  m real matrices by Rnm ; so that
A 2 Rnm means that A is an n  m matrix with real entries. We
write
A = (ai;j ) i=1;2;::;n 2 Rnm
j=1;2;::;m

DeÖnition 16 "Two matrices A and B are equal to ech other" means

that they are of the same size and they contain the same number at the
same adress that is
 
1 1 0
Example 17 A = is a 23 real matrix. Its entries are
1 5 12
integers .Its (2; 3)-entry (its entry in the second row and third column
) is 12: We have a1;1 = a1;2 = a2;1 = 1: What does this mean?
2 3
11 2 5
Example 18 B = 43=2 4=3 3=55 is a square matrix of size 3
0 1 1
.that is the number of rows of this matrix is equal to the number of
columns;/it is a 3  3-matrix.
 
1 1 0
Example 19 The transpose of the matrix A = is
1 5 12
33
34 Matrices
2 3 2 3
1 1 11 3=2 0
AT = 41 55 and the transpose of B is B T = 42 4=3 1 5
0 12   5 3=5 1
3 5
The matrix C = is symmetric since C = C T :
5 7

Matrix Operations
DeÖnition 20 We can add two real matrices only if they are of the
same size : If A = (ai;j ) and B = (bi;j ) are both real matrices of the
same size n  M then A + B is again a matrix of the same size n  m
with the entry in the i-th row and j-th column equal to ai;j + bi;j: Thus

(ai;j ) + (bi;j ) = (ai;j + bi;j ):

Example 21 Find the sum of the following matrices

   
1 1 0 3 5
1. + this sum is not deÖned, the matrices
1 5 12 5 7
are not of the same size
     
1 1 0 2 3 4 1+2 1 + (3) 0 + (4)
2. + = =
1 5 12  1 5 12 1 + (1) 5 + 5 12 + 12
3 2 4
2 0 24

DeÖnition 22 We can multiply a matrix by a number:

c(ai;j ) = (cai;j ):
    
3 5 33 3  (5) 9 15
Example 23 3 = =
2 7 3  (2) 37 6 21

DeÖnition 24 We can multiply two matrices A and B in this order,

if the number of columns of the Örst factor is equal to the number of
rows of the second, that is, if A 2 Rrm and B 2 Rms then we can
speak of the product A  B; it is deÖned as the r  s real matrix whixh
has in the i-th row and j-th column the entry

ai;1 b1;j + ai;2 b2;j +    + ai ;m bm;j :

Matrices 35

If we denote the i-th row 2of A 3by A(i) = [ai;1 ; ai;2 ; :::; ai ;m ] and the j-th
b1;j
6 b2;j 7
6 7
column of B by B (j) = 6 .. 7 then the above deÖnition of multipli-
4 . 5
bm;j
cation of matrices shows that we can multiply A(i) with B (j) and the
result is a 1  1-matrix and hence can be identiÖed with its entry. Then
we can write

A(i)  B (j) = ai;1 b1;j + ai;2 b2;j +    + ai ;m bm;j

and hence
A  B = (A(i)  B (j) )i=1;2;::;r ;j=1;2;::;s :
   
3 5 1 1 0 3  1 + (5)  (1) 3  1 + (5)  (
Example 25 1. =
 2
 7 1 5 12 (2)  1 + 7  (1) (2)  1 + 7  (
8 28 60
9 37 84
2 3
  1 1    
1 1 0 4 5 1+1 1  5 2 6
2. 1 5 = = and
1 5 2 1  5 1 + 25 + 2 6 28
2 3 0 2 2 3
1 1   2 6 2
41 55 1 1 0
=4 6 26 105
1 5 2
0 2 2 10 4
      
1 2 2 4 0 0 6 12
3. = 6=
2 4 1 2 0 0 3 6
2 3
A(1)
6A(2) 7
6 7
4. If A = 6 .. 7 = [A(1) ; A(2) ; :::; A(m) ] and B = [B (1) ; B (2) ; ::; B (s) ] =
4 . 5
A
2 3 (r) 2 3
B(1) A(1) B
6 B(2) 7 6A(2) B 7
6 7 6 7
6 .. 7 Then AB = [AB (1) ; AB (2) ; ::; AB (s) ] = 6 .. 7 and
4 . 5 4 . 5
B(m) A(r) B

(AB)(k) = A  B (k) = b1;k A(1) + b2;k A(2) +    + bm;k A(m)

and

(AB)(k) = A(k) B = ak;1 B(1) + ak;2 B(2) +    + ak;m B(m)

36 Matrices

Properties of these Matrix operations

1. Matrix addition is associative: (A + B) + C = A + (B + C) for
any three matrices A; B; C of the same size

2. Matrix addition is commutative: A + B = B + A:

3. There exists a unique matrix O of any Öxed size so that A + O =

A for any matrix A of that size.

4. Every matrix A has an additive inverse , denoted by A so that

A + (A) = O:

5. c(A + B) = (cA) + (cB)

6. (c + d)A = cA + dA

7. (cd)A = c(dA)

8. Matrix Multiplication is associative (AB)C = A(BC)

9. Matrix Multiplication is not commutative.

10. There is an identity matrix In is the n  n-matrix with all the

entries o with thre exception of the entries on the maim diagonal
of this matrix which are all 1 so that In A = A and BIn = B:

11. Multiplication is distributive over Addition A(B + C) = AB +

AC and (B + C)A = BA + CA

12. (cA)B = A(cB) = c(AB)

DeÖnition 26 A square matrix A 2 Rnn is said to be invertible if

there exists a matrix B such that AB = BA = In : If A is invertible
then there exists only one matrix B such that AB = BA = In : That
matrix is denoted by A1 and is called the inverse of A:Clearly if A is
invertible then its inverse A1 is also invertible and (A1 )1 = A: And
if A and B are invertible matrices of the same size then AB is also
invertible and (AB)1 = B 1 A1 :
 
2 3
Example 27 1. A = Then A is invetible : if we let B =
   5 8
      
8 3 2 3 8 3 1 0 8 3 2 3
then = = :
5 2 5 8 5 2 0 1 5 2 5 8
   
a b 1 1 d b
2. A = is invertible if  = adbc 6= 0 and A = 
c d       c
 a
a b 1 d b d b a b
because ( ) = I2 = ( 1 ) :
c d  c a c a c d
Matrices 37

3. Solve the following system

x + 4y = 5
3x + 5y = 2

This system can be written in matrix form as the matrix equation

AX = B
 
   
1 4 x 5
where A = ;X= ;B= : So we get
3 5 y 2

AX = B ) A1 (AX) = A1 B ) (A1 A)X = I2 X = X = A1 B

Therefore we get the solution of the given system easily if the

coe¢cient matrix A of the system is invertible. In our case
  A
5 4
 17 x
is invertible with A1 = 17 3 1 So the solution is =
 33    17 17
y
17 = 1
33
13 17 13
:
17

Row Reduction and Row Reduced

Echelon Matrices
DeÖnition 28 A matrix A = (ai;j ) 2 Rnm is said to be a row re-
duced echelon matrix if the following conditions are satisÖed:

1. There exists a number r 2 f0; 1; 2; :::; ng such that the Örst r rows
of A are nonzero rows and the last n  r rows are all zero rows.

2. Suppose that the Örst nonzero entry of the j-th row appears in the
column kj : Then we have

(a) aj;kj = 1; (the Örst nonzero entry of the j-th row is 1)j
(b) ai;kj = 0 for all j 6= i 2 f1; 2; ::; ng (all the other entries of
the column containing the Örst nonzero entry of a nonzero
row are all zero)
(c) 1  k1 < k2 < k3 <    < kr1 < kr  m:

Example
  29 The following
  matrices
 are row reduced echelon matrices
0 0 0 1 3 1 4 0
; ;
0 0 0 0 0 0 0 1
38 Matrices

Example 30 The following matrices are not row reduced echelon. Say
which
 condition
  of the
 deÖnition is
 violated.
 
0 0 0 2 3 1 4 2 0 1 3
; ; ;
1 0 0 0 0 0 0 1 1 0 2

Theorem 31 For any matrix A 2 Rnm there exists an invertible ma-

trix P and a unique row reduced echelon matrix R such that P A = R:
This R is called the row reduced echelon form of A:

The row reduced echelon form R of A can be found by successive ap-

plication of elementary row operations to A. These are of the following
forms:

1. Interchange two rows the i-th and j-th rows of A denoted by

Ri $ Rj

2. Multiply a row by some nonzero number cRi ! Ri ; multiply the

i-th row by c and take the result as the i-th row,

3. Add a multiple of some row to another row : cRi + Rj ! Rj

multiply the i-th row by c and add it to the row j and take the
result as the j-th row.

   
3 6 7 3 6 7
Example 32 Let A = ! we have applied
  2 4 6  1 2 3
1 3 6 7 1 2 3
R ! R2 ;
2 2
! we have applied R1 $ R2 ;
 1 2 3
 3  6 7
1 2 3 1 2 3
! we have applied (3)R1 + R2 ! R2 ;
 3 6 7  0 0  16  
1 2 3 1 2 3 1 1 2 3
! we have applied 16 R2 ! R2 ; !
0 0 16  0 0 1 0 0 1
1 2 0
we have applied 3R2 + R1 ! R1 :
0 0 1  
1 2 0
Indeed = R is the row reduced echelon form of A: It is a
0 0 1  
1 6 7
row reduced echelon matrix ,P = 32 is invertible and P A = R:
2 3
2 3
1 2 3 0
Example 33 Let A = 42 4 5 1 5 an invertible matrix P
5 10 14 1
and the row reduced echelon matrix R so that P A = R:
We proceed as follows: we augment the given matrix on the r¨ght
hand side by the identity matrix we get [AjI3 ] and apply sucsessive
Matrices 39

elementary operations until we get on the left part an row reduced

echelon matrix [RjP ] Then P is the 2 matrix one has to mult¨ply
3 A
1 2 3 0 1 0 0
to get the row reduced echelon form.42 4 5 1 0 1 05 !
2 3 2 5 10 14 1 03 0 1
1 2 3 0 1 0 0 1 2 3 0 1 0 0
40 0 1 1 2 1 05 ! 40 0 1 1 2 1 05 !
0 2 0 1 1 5 0 1 3 0 0 0 20 3 1 1 3
1 2 0 3 5 3 0 1 2 3 0
40 0 1 1 2 1 05 :T hen R = 40 0 1 15 and P =
2 0 0 30 0 3 1 12 3 20 0 0 0 3
5 3 0 5 3 0 1 2 3 0
42 1 05 We have P A = 4 2 1 05 42 4 5 1 5 =
23 1 1 3 3 1 1 5 10 14 1
1 2 0 3
40 0 1 1 5 = R:
0 0 0 0

Example 34 A square matrix is invertible if and only if its row re-

duced
2 echelon
3 form is the identity matrix. using2this fact show that P 3
=
5 3 0 5 3 0 1 0 0
42 1 05is invertible and Önd its inverse. 4 2 1 0 0 1 05 ;
3 1 1 2 3 1 3 1 0 0 1
1 0 0 1 3 0
The row reduced echelon form is: 40 1 0 2 5 05 : And this
0 0 1 5 14 1
shows that the row reduced echelon2form of P3is I3 and hence P is
1 3 0
invertible and its inverse is P 1 = 42 5 05 :
2 32 53 14 2 1 3
1 3 0 5 3 0 1 0 0
Check: 42 5 05 4 2 1 05 = 40 1 05 :
5 14 1 3 1 1 0 0 1
40 Matrices
 Chapter 5

Linear Systems of Equations

and Their Solutions

A linear equation in n u nknowns x1 ;x2 ; :::; xn is an equation of the

form
a1 x 1 + a2 x 2 +    + an x n = b
with the real constants a1 ; a2 ; :::; an as coe¢cients and b as another real
constant.
A system of linear equations or shortly a linear system in n un-
knowns x1 ; x2 ; :::; xn consisting of m linear equations can always be
written in the standard form
a1;1 x1 + a1;2 x2 +    + a1;n xn = b1
a2;1 x1 + a2;2 x2 +    + a2;n xn = b2
 = 
am;1 x1 + am;2 x2 +    + am;n xn = bm
with the real number ai;j as the coe¢cient of the j-th unknown xj
in the i-th equation of the system and the real number bi being the
constant in the i-th equation .
We call A = ( ai;j ) 2 Rmn the coe¢cient matrix of the system,
the column matrix B = (bi ) 2 Rm1 the matrix of constants and the
column matrix X = (xj )j=1;2;::;n the matrix of unknowns. Then we can
write the system shortly as a matrix equation
AX = B:
41
42 Linear Systems of Equations and Their Solutions

The column matrix C 2 Rn1 is said to be a solution of the system if

the matrix AC of numbers is equal to the matrix B:

Example 35 1. 3x + 2y  5z = 122 is3a linear equation in 2

the vari-
3
5 6
ables (unknowns) x; y; z: C1 = 465 is a solution.C2 = 4 0 5 is
2 3 3 6
1
another solution. D = 415 is not a solution.
1

2. The following is a linear system in the unknowns x; y; z consisting

of two equations

x + 3y + 2z = 6
2x  3y + z = 0:
 
1 3 2
The coe¢cient matrix is A = ; the matrix of con-
2 3 1 2 3
  x
6
stants is B = and the matrix of unknowns is X = 4y 5 :We
0
  z
x + 3y + 2z = 6
have that the system and the matrix equation
2 3 2x  3y + z = 0
  x  
1 3 2 4 5 6
y = are equivalent.
2 3 1 0
z
2 3 2 3
2   2  
1 3 2 4 0 5= 6 :
C = 4 0 5 is a solution since
2 3 1 0
4 4
2 3 2 3
13   13  
1 3 2 4 5 5= 6 :
D = 4 5 5 is another solution:
2 3 1 0
11 11

3. The system

2x + 3y = 1
4x + 6y = 5
 
u
has no solution because for any C = 2 R21 we have AC =
      v   
2 3 u 2u + 3v 2u + 3v 1
= = cannot be equal to
4 6 v 4u + 6v 2(2u + 3v) 5
for any choice of u and v .
Linear Systems of Equations and Their Solutions 43

How do we decide whether a given system is con-

sistent, that is has a solution or not.
We write the system in standard form with a Öxed ordering (which is
actually arbitrary) of the unknowns and of the equations and determine
the coe¢cient matrix A and the matrix of constants B and write the
augmented matrix [AjB] of the system, for example : Given the
equations

3a  2 + 5b = c
5 = b + a
b + 3c = 4 + 2a

we write them in standard form taking the unknowns a; b; c in the usual

alphabetic order and the following ordering of the equations

ab = 5
2a + b + 3c = 4
3a + 5b  c = 2:

Then the augmented matrix of this system is

2 3
1 1 0 5
4
[AjB] = 2 1 3 45
3 5 1 2

Then we row reduce the augmented matrix by successive application of

suitable row operations ,that is we Önd an invertible matrix P such that
P [AjB] = [P AjP B] is a row reduced echelon matrix. Then P A = R
is the row reduced echelon form of A:
If the number of nonzero rows of R is less than the number of
nonzero rows of [P AjP B] then the system has no solution, because
then the matrix [P AjP B] has a row of the form (0; 0; :::0; 1) 2 R1(n+1)
which corresponds to an equation

0x1 + 0x2 +    + 0xn = 1

which has clearly no solution. Using the following Fact we can conclude
that the system with the augmented matrix [AjB] has no solution.
Fact The system with the augmented matrix [AjB] and the system
with the augmented matrix [P AjP B] have exactly the same set of
solutions if P is an invertible matrix.
So the system is consistent if and only if the rank of A is equal to
the rank of [AjB]: Here rank of a matrix is deÖned as the number of
nonzero rows of its row reduced echelon form.
44 Linear Systems of Equations and Their Solutions
2 3
1 1 2
Example 36 41 1 25 has rank 1 because its row reduced echelon
2 33 3 6
1 1 2
form is 40 0 05 and has only one nonzero row.
0 0 0
2 3
1 1 3 0
Example 37 The system with the augmented matrix 42 1 0 45
3 0 3 5
is
2 inconsistent
3 because the row reduced echelon form of this matrix is
1 0 1 0
40 1 2 05 The rank of the coe¢cient matrix is 2 but the rank of
0 0 0 1
the augmented matrix is 3.So the system has no solution.

How can we Önd all solutions if we know the system

is consistent?
Let P be an invertible matrix so that [P AjP B] is the row reduced
echelon form of the augmented matrix [AjB] of the given system, and
assume that the rank of A is equal to the rank of [AjB] and hence the
system has solutions.
If r = the rank of A = the number of variables = m = the number
of columns of A then the row reduced echelon form of [AjB] is equal to
 
Ir C
[P AjP B] =
O1 O2
where O1 is the (n  r)  r-Zero matrix and O2 is the (n  r)  1-Zero
matrix C is an r  1-matrix. Then the system has a unique solution;
namely X = C:
If r < m then we choose m  r of the unknowns as free parameters
which can take any real number as value and determine the remaining r
basic variables from the r equations corresponding to the nonzero rows
of [AjB]: The basic variables xkj will be the variables corresponding to
the Örst nonzero entries of the nonzero rows of [AjB]; j = 1; 2; :::; r:
Observe that the variable xkj occurs with nonzero coe¢cient only in
the j-th equation. You will follow the details in the examples.
2 3
1 1 3 0
Example 38 Consider the system with the augmented matrix 42 1 0 45
2 3 0 33 4
1 0 1 34
The row reduced echelon form of the augmented matrix is 40 1 2 43 5
0 0 0 0
Linear Systems of Equations and Their Solutions 45

We have a consistent system wirh r = 2 < 3 = m: The Örst two

variables x1 and x2 are basic variables and x3 = t will be chosen as
parameter.The equations corresponding to the nonzero rows of the aug-
mented matrix are then
4
x1 + x3 =
3
4
x2  2x3 =
3
from which we obtain
4
x1 = t +
3
4
x2 = 2t +
3
x3 = t
2 3
t + 43
and hence the set of all solutions is S = f4 2t + 43 5 : t 2 Rg:
t
Example 39 Solve the following system (Find the set of all solutions
of the system):
2x + 3y = 9
3x  5x = 10
6x + 11y = 13
x + 2y = 1:
2 3
2 3 9
63 5 10 7
The augmented matrix of the system is 6 7
4 6 11 135 and has
1 2 1
2 3
1 0 15
60 1 7 7
the row reduced echelon form 6
40
7 : So we have a unique so-
0 0 5
  0 0 0
15
lution, namely :
7
Example 40 Solve the following system (Find the set of all solutions
of the system):
2x1 + x2  x3 + 2x4 = 0
2x1 + x2 + x3 + 3x4 = 1
4x1 + 2x2 + 5x4 = 1
2x1  x2  3x3  4x4 = 2
46 Linear Systems of Equations and Their Solutions
2 3
2 1 1 2 0
62 1 1 3 17
The augmented matrix of this system is 6 7,
44 2 0 5 15
2 1 3 4 2
2 1 3
1 2 0 54  14
60 0 1 12  12 7
and its row reduced echelon form is 640 0
7 : The system is
0 0 0 5
0 0 0 0 0
consistent and is equivalent to the system
1 5 1
x1 + x2 + x4 = 
2 4 4
1 1
x3 + x4 = 
2 2
and we see that x1 and x3 are basic variables and can be expressed using
these equations linearly in terms of the other variables x2 and x4 which
can be taken as free parameters ranging independently and freely over
R: The set of solutions is then
2 1 1 3
 4  2 u1  54 u2
6 u1 7
S = f6 7
4   u2 5 : u1 ; u2 2 Rg
1 1
2 2
u2
 Chapter 6

Di§erentiation

y y  y0
Let y = f (x) be a given function . The quantity = =
x x  x0
f (x)  f (x0 )
is called the di§erence quotient of f at the point
x  x0
(x0 ; y0 ) = (x0 ; f (x0 )): For example the di§erence quotient of

y x2  x20 (x + x0 )(x  x0 )
1. y = x2 at (x0 ; x20 ) is then = = =
x x  x0 x  x0
x + x0 ;
y x3  x30 (x2 + xx0 + x20 )(x  x0 )
2. y = x3 at (x0 ; x30 ) is then = = =
x x  x0 x  x0
x2 + xx0 + x20 ;
2 2 1 1 1 1
2 y
2 x 3  x03 (x 3 + x03 )(x 3  x03 )
3. y = x at (x0 ; x0 ) is then
3 3
= = 2 1 2 1 =
x x  x0 3
1
3 3 3
1
(x + x x0 + x0 )(x  x0 )
3 3

1 1
(x 3 + x03 )
2 1 1 2 :
(x 3 + x 3 x03 + x03 )

In all these cases the di§erence quotient considered as a function

of x is continuous at x0 (in case 3 if x0 6= 0) so that its value at x0
y
coincides with the value of limx!x0 and we have respectively that
x
y
1. limx!x0 = x0 + x0 = 2x0 ;
x
47
48 Di§erentiation

y
2. limx!x0 = x20 + x0 x0 + x20 = 3x20
x
1 1 1
y (x 3 + x 3 ) 2x 3 2 1
3. limx!x0 = 2 0 1 1 0 2 = 02 = x0 3
x (x03 + x03 x03 + x03 ) 3x03 3
for any x0 :

dy
We deÖne the derivative (x0 ) of y = f (x) at the point
dx
(x0 ; f (x0 )) as
dy y
(x0 ) = lim
dx x!x0 x

and put
dy y
(x0 ) = (x0 )
dx x
y
whenever (considered as a function of x) is a continuous function
x
at x0 : One can show that if

dy
y = xa then (x) = axa1 for any a 2 Q and x an inner point of the domain :
dx
dy
We denote the derivative of y = f (x) at the point x by (x) or
dx
df
(x) or y 0 (x) or f 0 (x): And the process of getting the derivative from
dx
the function is called the di§erentiation.

Example 41 Find the derivative of the following functions at the in-

dicated points

1. y = 3x3 at x0 = 2:
p
2. y = 3
x at x0 = 8:

Example 42 Di§erentiate the following functions (Önd the derivative

functions of the following functions)

1. y = x5 ;
p
5
2. y = x3 :
Di§erentiation 49

Di§erentiation Rules
Di§erentiation has the following properties:

1. If f (x) = c is the constant function with the value c then f 0 (x) =

0 for any x:

2. If y = cf (x) then y 0 = cf 0 (x):

3. If y = f (x) + g(x) then y 0 = f 0 (x) + g 0 (x):

4. If y = f (x)g(x) then y 0 = f 0 (x)g(x) + f (x)g 0 (x):

f (x) f 0 (x)g(x)  f (x)g 0 (x)

5. If y = then y 0 =
g(x) g(x)2

6. If y = f (g(x)) then y 0 = f 0 (g(x))g 0 (x):

Example 43 Find the derivative functions of the following functions

1. y = 3x + x3
p
2. y = x3 + x1
p
3. y = (x + x3 )( x3 + 5x1 )

x2 + 3x
4. y =
x2  3x
r
x2 + 3
5. y =
x2  3
p
5
6. y = 1 + 2x + x3

1
7. y = p
x + x2 + 1
50 Di§erentiation

Applications of the derivative

1. Approximation One can approximate arbitrary highly di§er-
entiable functions by polynomial functions: if f (x) is n times
di§erentiable at the point x0 then there is an open interval (a; b)
containing x0 so that f (x) is approximately equal to the polyno-
mial

1 1
f (x0 )+f 0 (x0 )(xx0 )+ f 00 (x0 )(xx0 )2 +  + f (n) (x0 )(xx0 )n
2! n!

Exercise 44 (a) Express the polynomial p(x) = 1+xx2 +2x3

in terms of powers of (x + 2):
p(x) = 1+xx2 +2x3 = p(2)+p0 (2)(x+2)+ 2!1 p00 (2)(x+
2)2 + 3!1 p000 (2)(x + 2)3 Thus we need the derivatives of p :
p0 = 1  2x + 6x2 ; p00 = 2 + 12x; p000 = 12 and p(2) =
1  2  4  16 = 21; p0 (2) = 1 + 4 + 24 = 29; p00 (2) =
2  24 = 26; p000 (2) = 12; hence

p(x) = 21 + 29(x + 2)  13(x + 2)2 + 2(x + 2)3

1
(b) Approximate the function y = e 2 x around x0 = 0 by poly-
nomials of degrees 1,2,3 and 4 using the fact that (ex )0 = ex
and e0 = 1:
1 1 1 1 12 x
We have y 0 = 12 e 2 x :y 00 = 14 e 2 x ; y 000 = 18 e 2 x and y (4) = 16
e
and hence we get

1
y1 = 1 + x
2
1 1 2
y2 = 1 + x + x
2 8
1 1 2 1 3
y3 = 1 + x + x + x
2 8 48
1 1 2 1 3 1 4
y4 = 1 + x + x + x + x
2 8 48 384
Di§erentiation 51

12
y
10

-5 -4 -3 -2 -1 1 2 3 4 5
x

1
In this Ögure red curve is y = e 2 x and the black curve is y4 :

2. The Örst degree approximating function y = f (x0 )+f 0 (x0 )(xx0 )

of the function y = f (x) at x0 is actually the tangent line to the
graph of y = f (x) at the point (x0 ; f (x0 )): This means that the
value f 0 (x0 ) of the derivative of a function y = f (x) at x0 is the
slope of the tangent line to the graph of y = f (x) at the point
(x0 ; f (x0 )):

Exercise 45 (a) p Find the equation of the tangent line to the

3 2
graph y = x + 4x + 3 at the point (1; f (1))
p 2
f (1) = 3 1 + 4 + 3 = 2 and f 0 (x) = 13 (x2 + 4x + 3) 3 (2x + 4)
6 1
and hence f 0 (1) = 3 p
3 2 = 2 : So the equation of the tangent
8
line to the graph of y = f (x) at the point (1; 2) is y  2 =
1
2
(x  1); that is
1 3
y = x+ :
2 2
(b) If possible Önd the lines which pass through (5; 11) and are
tangent to the curve y = x2 :
The point (5; 11) is not on the curve 11 = 52 is not true.
Suppose that there exists a line passing through (5; 11) and
is tangent to the curve y = x2 at the point (x0 ; x20 ) Then the
line passing through these two di§erent points has the slope

11  x20
:
5  x0
52 Di§erentiation

On the other hand this line is tangent to the curve at (x0 ; x20 )
and hence slope equal to the derivative of the curve at x0 : So
we get
11  x20
= 2x0
5  x0
and hence 10x0  2x20 = 11  x20 ; that is

x20  10x0  11 = 0
p
2 :+44 p
This quadratic equation has the roots 10 10
2
= 5  36
Thus x0 = 11 or x0 = 1. So we have two such tangent
lines
y  121
= 2  11
x  11
y1
= 2  (1)
x+1
that is

y = 22x  121
y = 2x  1

3. Curve sketching. For curve sketching the ínformation one ob-

tains from the local behaviour of the function and hence from
the derivative function are very valuable For example we know
that
Fact If the derivative g 0 of a function g is positive (respectively
negative) on the interval I then g is an increasing (respectively
decreasing) function on I:
Fact If the second derivative g 00 of a function g is positive (respec-
tively negative) on the interval I then the graph of g is concave
up (respectively concave down)

(x2)2
Exercise 46 Sketch the graph of the following curves y = x2 +1
and y = (x+1)(x3)
x2 4
:

4. Extreme values
Fact A continuous function deÖned on a closed and bounded
interval [a; b] (with a; b 2 R and a < b) has a maximum and a
minimum, that is there exist points x1 and x2 in [a; b] so that

f (x1 )  f (x)  f (x2 )

Di§erentiation 53

for any x 2 [a; b]:

Remark If the function is not continuous or the region where
it is deÖned is not bounded or is not closed then extreme values
do not need to exist.
Remark If f : [a; b] ! R is continuous and if f is di§erentiable
on the interval (a; b) then the exteme values occur either on the
boundary points x = a or x = b or at some inner point x 2 (a; b)
where the derivative vanishes , that is f 0 (x) = 0:

Exercise 47 Find the extreme values of y = 11x3 + 15x2  3x

on [2; 1]:
p
0 2 0 10 102 411(1)
We have y = 33x +30x3 .Thus y = 0 if x = 211
=
56 1
11
that is x = 1 or x = 11
: Both are in the interval . So
1
the extreme values may occur only at the points 2; 1; 11 ; 1.The
17
values of the function at these points are 22,7,  121 . 23 So the
maximum is 23 and occurs at the point x = 1 and the minimum
is 22 and occurs at 2:

Exercise 48